Pile Load Capacity 4

[ONENGINEER]

In designing the bearing capacity of a 10" pile penetrating into sand,I have used 5 different methods. Using Qa = 1/3(Qb + Qs) I have obtained Qa values ranging from 25 ton (Decourt) 36 ton (Coyle, ..), 38 ton (Janbo Mayerholf), 80 ton (Mayerholf) and 152 ton (betta method). One would like to compare the results to find the most safe reasonable ones. However the range is too wide. I am writing to hear your possible comments, Thank you.

 

[Kpaudel]

Dear onengineer for detailed comment you have to post some physical properties of these sandy layer, including compactness of strata, size of sand grains (seive analysis parameter), then u can goes as nearly as reality...

 

[jgailla]

There is a great article by Peck (if I remember correctly?) demonstrating that every pile capacity is 80 tons. Of course completely tongue in cheek.
I don't think sieve analysis is necessarily relevant here.
Pile design capacity can be reasonably approximated in sand using N-values, confirmed in the field using static load tests or PDA.
I would expect a 10" pile driven 60 or 80 feet into loose down into medium dense sand to be about 80 to 100 tons assuming a whole bunch of stuff.
What method is most commonly used in your area?
The strata will have a huge affect on the capacity. With what you've given, any of the numbers could be reasonable.
From 25 to 152 tons seems to be too large of a range to be reasonable and I assume you may be comparing methods with different assumptions of soil types which are not applicable to your situation.

 

[BigH]

Take a look at comparisons of Nq for piles by various authors and I fully understand why one can "get" 52 to 152 tons. You will find in some textbooks curves of Nq from various authors. For example with phi = 35deg, DeBeer's Nq value is about 600, Terzaghi is about 70; Berezantev is about 90; Vesic is about 75; Meyerhof (driven) is about 300. So, what does one use? Most suggest use of Berezantev as the "most reliable". But, in the end, it comes down to judgment for design and then use proper methods in construction to confirm design (and, perhaps, refine the design values).

 

[Kpaudel]

As sieve analysis along with SPT N value and laboratory shear parameter, you can apporach nearly phi value, so u can easily decide the methods appropriate for your site.

 

[ONENGINEER]

jgailla. The piles are intended to be closed end steel pipe piles filled with concrete after installation and are intended to penetrate into medium to dense sand up to about 5 m (15 ft). The sand is overlain by about 20 m (60 ft) of clayey silt. For the pile load capacity comparisons, any contribution from the silt materials have been ignored. Mayerholf method gives about 100 tons (this is 1000 kN). Decourt gives the lowest value of 23 ton (230 kN). As BigH suggested this is to some extent due to variations in Nq as well as the fact that some (which give higher values for pile capacity) take into account the effective stresses directly in their formulations. The pile loading test is after the installation so I am trying to predict the pile capacity as close as to the real value as I can at this stage.

 

[jgailla]

Are you adding any fill to the site? Could negative skin friction develop in a compressible silt layer? This could make your pile capacity effectively nothing.
It comes down to what is typically used in your area with its specific soil types and strata. Different approximations work better in different areas after validation with field testing. Drop back and look at how the equations were developed and in what soil types. One equation may work well when phi is low while another is better when phi is high.
There should be someone supervisory in your office who could give you some guidance on which model has worked best for your area.
Are there older calculations in your office you could look at?
From the limited information you've given, I would be leaning more to the low end of the estimates, but that's just a gut response. If you don't have enough experience in this area to choose you need to find someone who does.
You'll either underestimate the pile capacity, costing lots of money and you won't get any more work, or overestimate the pile capacity, causing change orders during construction costing lots of money and you won't get any more work.
The other thing to think about is the size and geometry of the building. The structural engineer will need a certain of piles just for grade beam intersections and spaced reasonably to avoid huge grade beam sizes. Perhaps you could discuss the loading with the structural engineer who may already have some pile location geometry worked out. It could be better to go with a slightly larger pile.
To your original question, though, no one on the internet will be able to tell you which is the right one to use.

BigH knows a lot more about this than me.

 

[motorbiketocrank]

How I'd handle this would depend largely on the timing of when the production piles need to be ordered in comparison to when the test pile/load test will be performed. Best case is if you can do the load test before production piles are ordered. Bring a longer pile than you think you need for the test. Run a wave equation analysis in advance and use PDA during installation. Then based on the PDA stop the pile at the desired depth, do the static test or CAPWAP and then order production piles based on the result. If production piles need to be ordered before the test, I'd generally order them on the long side. Based on your post, you already plan on going at least 75' so adding another 10' per pile on the order wouldn't add too much on the % length/cost and would provide a buffer and also account for variability around the site. So if you end up cutting off most piles not too big of a deal.....and would provide material for select areas that need extra length to be spliced on.

 

[motorbiketocrank]

Should also have mentioned that I agree with the other posts, especially the question about negative skin friction. You'll need to think through how you'll subtract the contribution the silts provide to capacity during the load test and also add their downdrag load as applicable depending on the relative settlement behavior of that layer as compared to allowable pile movement.

 

[ONENGINEER]

I am really grateful for the good advices here. However, I am really looking for the assessment of various bearing capacity methodologies for a single pile penetrating into sand, which I think is one of the basics of foundation engineering. I am sure experienced engineers and university professors have come accross this issue many times.

 

[jgailla]

ONENGINEER,
You are going to get more practical than theoretical help on this website.
Good luck with your learning.

 

[VAD]

ONENGINEER:

Pile design by static analysis is an ever revolving topic and numerous articles have been written on the subject. Traditionally the relationship you have used shaft resistance plus toe resistance divided by a Factor of Safety in working stress design (WSD) is often accepted to provide somewhat reasonable settlements. Meyerhof recommends FOS of 4 whan SPT tests are used. Meeting a prescribed settlement is often the crux of a pile design and not necessarily the factored resistance (LRFD) or allowable resistance (WSD). Note that the methods that you have used are based on the failure resistance of the pile, the value of which has been reduced to a somewhat safe value by your FOS. This of course is at best a guess or is believable based on past experience.

Very often when we are asked to provide a factored resistance or pile allowable resistance we do so without checking the settlement that the pile will undergo. Very often we invoke statements if we are up to it of settlement will be less than an inch etc. There are approaches which can allow us to calculate the settlement and hence give us a better appreciation for the load or resistance to be recommended. Although taken as a simple exercise in calculating shaft and toe resistances there is a lot more complexity to the subject which undfortunately was not often taught at all Universities in the older days. However, this is gradually changing. There are two experts Poulos and Fellenius who have made significant contributions to the subject and their papers are worthwhile reading also Randolph and Flemings text illustrate how the settlement of a pile can be determined. We have not touched on load transfer which further determines whether design should be based on shaft resistance or toe resistance only or a combination of both. Remember settlement again will govern. Note that the structural loads may or may not be provided, but it is always best to obtain these loads when providing design recommendations. This is also a debatable subject

Just to indicate that I have noted there is still the misunderstanding of downdrag or regative skin friction. Since our conventional pile design is based on ultimate resistance failure the negative skin friction has no part to play in its ultimate resistance determination. The drag load that results from the downdrag is important in determining the structural integrity of the pile. This dragload is also used in settlement determination. The reason why the drag load does not play a part in the ultimate resistanece is that in a failure mode the negative friction is eliminated as the pile plunges causing a reversal of the direction in the negative friction. This has been the subject of much debate over the years, but if you think about it the concept makes perfect sense.